Skip to main content
The OpenShift Container Platform monitoring stack ships its own Alertmanager (alertmanager-main in the openshift-monitoring project). The official documentation lists PagerDuty, Webhook, Email, Slack and Microsoft Teams as receiver types, and states that the features of a supported upstream Alertmanager version are supported in OpenShift as well. The webhook receiver sends Alertmanager JSON to the URL you set (version is 4; each alert carries status, labels, annotations, startsAt, endsAt, generatorURL, fingerprint). The Flashduty Prometheus integration accepts this format, so no separate OpenShift integration is needed: create a Prometheus integration in Flashduty and configure its push URL as a webhook receiver in the OpenShift Alertmanager.

​
In Flashduty On-call

Get an integration push URL in either of the two ways below. Choose the Prometheus integration type in both, not Chronosphere.

​
Use a dedicated integration

  1. In the Flashduty console, go to Channels and open a channel
  2. Go to Settings → Integrations → Dedicated integrations and click Add an integration
  3. Select Prometheus and click Save
  4. Open the generated integration card and copy the Push URL, in the form https://api.flashcat.cloud/event/push/alert/prometheus?integration_key=<integration key>

​
Use a shared integration

  1. In the Flashduty console, go to Integration Center → Alert Events
  2. Select Prometheus and enter an integration name
  3. Configure the default route and select a channel; you can add more rules under Routes after creation
  4. Click Save and copy the generated Push URL

​
Configure in OpenShift

The OpenShift documentation offers two ways to configure a receiver; use either one. Both require the cluster-admin role.
1

Option 1: the web console

  1. In the Administrator perspective, go to Administration → Cluster Settings → Configuration → Alertmanager
  2. In the Receivers section, click Create Receiver
  3. Enter a Receiver name (for example flashduty) and select Webhook as the Receiver type
  4. In the webhook endpoint field, enter the full push URL copied above, including ?integration_key=...
  5. Show advanced configuration has the option to send resolved alerts; it is on by default, leave it unchanged
  6. In Routing labels, use Add label to add label names and values that select which alerts go to this receiver; per the documentation, firing alerts whose labels match all selectors are sent to the receiver, and adding labels here makes the label values match exactly
  7. Click Create
2

Option 2: edit the alertmanager-main secret

  1. Export the current configuration:
  2. Add a receiver and a route to alertmanager.yaml. The example adds a flashduty receiver to the default configuration and routes alerts whose severity is critical or warning to it. send_resolved defaults to true in the Alertmanager webhook configuration; it is written out to make sure resolved notifications stay on:
Replace <integration key> with your Flashduty integration key. Keep the new route after the Watchdog route: Watchdog is an alert that fires continuously, and the OpenShift documentation describes it as a check on the notification path, not a fault alert to forward to Flashduty.
  1. Replace the secret with the edited file:
3

Verify

Trigger a cluster alert (the OpenShift documentation points to the Red Hat Customer Portal article “Send test alerts to Alertmanager in OpenShift 4”) and confirm Flashduty receives an active alert; after it resolves, confirm the original alert closes in Flashduty.

​
Field mapping

One notification can carry several alerts, and Flashduty handles each one separately (up to 100 per request).

​
Recovery and deduplication

  • send_resolved defaults to true for the Alertmanager webhook receiver. On recovery it sends a notification with status set to resolved, and Flashduty closes the alert with the same fingerprint.
  • If you turn resolved notifications off under Show advanced configuration (or set send_resolved: false in YAML), Flashduty receives no recovery requests; enable auto-resolve timeout on the channel.
  • Alertmanager repeats notifications for alerts that are still firing, per repeat_interval; Flashduty deduplicates by fingerprint, so no duplicate alerts are created.

​
Troubleshooting

  • Flashduty returns Invalid parameters: the push URL is incomplete, integration_key is missing, or the integration type is not Prometheus
  • Alerts do not resolve: confirm the receiver does not set send_resolved: false
  • No alerts arrive: confirm the route matchers or Routing labels match the alert labels and the cluster can reach api.flashcat.cloud (with a cluster-wide HTTP proxy, see the proxy_from_environment note in the OpenShift documentation); oc exec alertmanager-main-0 -n openshift-monitoring -- amtool config routes show --alertmanager.url http://localhost:9093 shows the effective routes